4814 cording to eq 32, within relatively narrow limits. Proton transfer from protonated dimethyl carbonate to Fe(C0)5 is readily observed, but no proton transfer is seen from NH4+. This is interpreted to mean that PA[(CH30)2CO] I PA[Fe(CO)s] IPA[NH3]. Recent results from our laboratory indicate that PA[(CH30)2CO] = 203 f 2 k ~ a l / r n o l and ,~~ thus PA[Fe(C0)5] is assigned as 204 f 3 kcal/mol. Iron pentacarbonyl is thus actually quite a strong base in the gas phase. The difference in gas phase and solution behavior is attributable to the understandably poor solvation of H Fe( C0)5+. The decomposition of HFe(C0)5+ by loss of CO is observed when it is formed in a sufficiently exothermic proton transfer reaction. This reaction is observed with all proton donors less basic than dimethyl ether. These results indicate that D[HFe(C0)4+-CO] = 23 f 10 kcal/mol, which is not significantly different from the C O binding energy in the parent ion (22 kcal/mol). Conc1usi ons The results described above indicate that ICR has tremendous potential for illuminating certain aspects of organometallic chemistry. It is to be stressed that such experiments are performed at low pressure in the gas phase and provide information relating to the intrinsic reactivity of the molecules studied in the absence of solvent effects. The results above suggest that aspects of particular interest may be: (1) formation of polynuclear metal clusters containing varying numbers and types of ligands, (2) ligand substitution processes, (3) determination of relative ligand binding energies, (4) accurate determination of transition metal basicity, (5) characterization of processes involving both electrophilic and nucleophilic attack on neutral metal complexes, (6) generation and study of unusual u- and s-bonded organometallic complexes, and (7) photochemistry of gaseous organometallic ions.31 References and Notes (1) Dreyfus Teacher-Scholar, 1971-1976. (2) F. Basoio and R. G. Pearson, "Mechanisms of inorganic Reactions", 1st
ed, Wiley. New York, N.Y.. 1967, Chapter 7, and recent references cited therein. (3) (a) J. L. Beauchamp, Annu. Rev. Phys. Chem., 22, 527 (1971); (b) G. A. Gray, Adv. Chem. Phys., I O , 141 (1971); (c) J. D. Baldeschwieler and S. S. Woodgate. Acc. Chem. Res., 4, 114(1971). (4) (a) R. H. Wyatt, D. Holtz, T. B. McMahon. and J. L. Beauchamp. horg. Chem., 13, 1511 (1971); (b) J. C. Haartz and D. H. McDaniel, J. Am. Chem. Soc., 05, 8562 (1973); (c) W. T. Huntress, Jr., and R. F. Plnizzotto, Jr.. J. Chem. Phys., 50, 4742 (1973). (5) M. S. Foster and J. L. Beauchamp, J. Am. Chem. Soc., 03, 4924 (1971). (6) R. C. Dunbar. J. F. Ennever, and J. P. Fackler, Jr., Inorg. Chem., 12, 2734 (1973). (7) R. D. Bach, J. Gauglhofer. and L. Kevan, J. Am. Chem. Soc., 94, 6860 (1972). (8) J. Muller and K. Fenderl, Chem. Ber., 103, 3141 (1970); ibid.. 104, 2199, 2207 (1971); J. Muller and W. all, Chem. Ber., 106, 1129 (1973); /bid.,107, 2084 (1974). (9) S. M. Schildcrout, J. Am. Chem. Soc., 95, 3846 (1973). (10) Trapped-ion techniques used in the present work are described in T. B. McMahon and J. L. Beauchamp, Rev. Sci. hstrum.. 43, 509 (1972). Absolute pressures were measured according to the procedure in R. J. Biint, T. B. McMahon, and J. L. Beauchamp, J. Am. Chem. Soc., 06, 1269 (1974). (1 1) G. Distefano, J. Res. Natl. Bur. Stand., Sect. A, 74, 233 (1970), and references cited thereln. (12) S. Pignataro, A. Foffani, F. Grasso, and B. Cantone, Z. Phys. Chem. (frankfurtam Mah), 47, 106 (1965). (13) J. L. Franklin, J. G. Dillard, H. M. Rosenstock, J. T. Herron, K. Draxl, and F. H. Field, k t l . Stand. Ref. Data Ser., Natl. Bur. Stand., No. 26 (1969). (14) For a description of the gas-phase ion chemistry of CHsF, see A. 0. Marshall and S. E. Buttriil, Jr., J. Chem. Phys., 52, 2752 (1970). (15) J. P. Coliman and W. R. Roper, Adv. Organomet. Chem., 7,53 (1968). (16) J. Haipern. Acc. Chem. Res., 3 , 386 (1970). (17) J. Lewisand B. F. 0. Johnson, Acc. Chem. Res., 1,245 (1968). (18) 8. F. G. Johnson, J. Lewis, i. G. Williams, and J. M. Wilson, J. Chem. Soc. A, 341 (1967). (19) C. A. Tolman, Chem. Soc. Rev., 1, 337 (1972). (20) F. A. Cotton and G. Wiikinson, "Advanced Inorganic Chemistry", 2nd ed, interscience. New York, N.Y., 1966. Chapter 27. (21) E. W. Abel, R. A. N. McLean, S. P. Tyfield, P. S. Braterman, A. P. Walker, and P. J. Hendra, J. Mol. Spectrosc.. 30, 29 (1969). (22) See, for example, (a) ref 22, p 746, and ref 2, p 538; (b) F. A. Cotton, lnorg. Chem., 3, 702 (1964): (c) W. A. G. Graham, bid., 7, 315 (1968). (23) E. L. Muetterties, Inorg. Chem., 4, 1841 (1965). (24) T. Kruck and M. Noack, Chem. Ber., 07, 1693 (1964). (25) B. F. G. Johnson and J. A. McCleverty. frogr. Inorg. Chem., 7, 277 (1966). (26) D. F. Shriver, Acc. Chem. Res., 3, 231 (1970). (27) J. C. Kotz and D. G. Pedrotty, Organomet. Chem. Rev., Sect. A, 4, 479 (1969). (28) Reference 2, p 545. (29) A. Davison. W. McFariane, L. Pratt, and G. Wilkinson, J. Chem. Soc., 3653 (1962). (30) R. R. Corderman, R. H. Staley, and J. L. Beauchamp. unpublished results. (31) R. C. Dunbar and B. 8. Hutchinson, J. Am. Chem. Soc., 06, 3816 (1974).
Ion-Molecule Reactions and Gas-Phase Basicity of Ferrocene M. S . Foster and J. L. Beauchamp*' Contribution No. SO07 from the Arthur Amos Noyes Laboratory of Chemical Physics. California Institute of Technology, Pasadena, California 91 125. Received October 31, 1974
Abstract: The gas-phase ion chemistry of ferrocene is investigated using the techniques of ion cyclotron resonance spectroscopy. Product distributions and rate constants for the principal primary ions are determined using trapped ion methods. Proton transfer reactions in mixtures of ferrocene with other molecules place fairly accurate limits on the gas-phase basicity of the molecule, leading to a proton affinity (213 i 4 kcal/mol) which is slightly less than methylamine. The gas-phase measurements are free from complicating solvent effects which have previously led to some unwarranted conclusions about transition-metal basicity.
The unusual structures and bonding of dicyclopentadienyl metal complexes (metallocenes) have stimulated many mass spectral investigations of these molecule^.^^^ Of particular interest is the recent high-pressure study of ferrocene Journal of the American Chemical Society
/
97:17
/
by Schildcrout? who demonstrated the occurrence of several ion-molecule reactions. The present paper describes an ion cyclotron resonance (ICR) study of the ion chemistry of ferrocene, both alone and in mixtures with several other
August 20, 1975
4815
molecules. This investigation continues our efforts to apply the powerful ICR technique to transition-metal organometallic specie^.^ Ferrocene is a logical choice for examination because of its high stability, adequate vapor pressure, and position as the most chemically significant metallocene. In addition to elucidating the reactions of the ions derived from ferrocene, these experiments place fairly accurate limits on the gas-phase basicity of the molecule. The latter information is particularly significant in view of the mechanism invoked for electrophilic substitution in the metallocenes.6 It is to be emphasized that the experiments described are performed a t low pressure in the gas phase, in the absence of complicating solvent effects which have apparently led to some unwarranted conclusions about transition-metal basicity. Experimental Section The general features and operating characteristics of ICR instrumentation have been previously described in detail.’ Equipped with a “flat” cell modified for trapped-ion operation,8 the spectrometer used in the present study was constructed at Caltech and is capable of improved resolution over a wider mass range (upper limit m/e 600) than previously available instruments. The pulse circuitry which permits both trapped-ion and normal drift-mode operation will be detailed e l ~ e w h e r e . ~ Ferrocene was obtained from Aldrich Chemical Co. and used as supplied; no impurities were evident in the ICR mass spectrum. The vapor pressure of ferrocene at room temperature is about Torr,I0 which was adequate for all of the present experiments when the vapor from a crystalline sample was introduced by means of the normal inlet system (i,e,, through a variable leak valve). Heating the sample was not required. All experiments were performed at ambient temperature (25-30’). Accurate pressure measurements were made with a Schulz-Phelps type ion gauge calibrated against a MKS Baratron Model 90H1-E capacitance manometer in the manner previously described.” The estimated uncertainty in absolute pressures, and thus in all rate constants reported, is *lo%. Ferrocene was extremely well-behaved in the ICR spectrometer and no experimental difficulties were encountered.
-
O.Ol0
+ Fe(C5H5)2
FeCsH5+
+ Fe(CsH5)2
FeC5H5+
ki +
k2
+
(1)
+ FeCsH5
(2)
Fe(C5H5)2+ Fe Fe(CsH5)2+
+ Fe(CsH5)2
Fez(CsHd3’
(3)
tem, the latter being responsible for the product a t m/e 307, which presumably has a “triple-decker sandwich” struct ~ r eThe . ~ parent ion appears unreactive, a conclusion verified by lowering the electron energy to 11 eV, which is below the appearance potentials of Fe+ and F ~ C S H ~ and noting that no reaction products occur out to m/e 500. Processes 1-3 are the same as those deduced by Schildcrout from high-pressure mass spectrometry experiments4 With no neutral product, process 3 is unusual for a n ion-molecule Foster, Beauchamp
100
150
200
250
300
Figure 1. Variation with time of positive ion abundances in ferrocene following a 10 msec, 70 eV electron beam pulse at a pressure of 4.3 X IO-’ Torr. Ij and mi represent the ICR single resonance intensity and is proportional to ion mass of the i t h ionic species. The quantity abundance.
kj +
50
TIME (msec)
Results and Discussion Ion Chemistry of Ferrocene. The 70 eV ICR mass spectrum of ferrocene at 1.0 X lo-’ Torr agrees with previous electron impact results.2 The ions observed and their relative abundances are: Fe(CsH5)2+ (loo), FeCsHs+ (48), and Fe+ (16). The parent ion was most prominent a t all electron energies. A t an electron energy of 70 eV, one ion-molecule reaction product appears a t m/e 307 as the pressure of ferrocene is raised. At a pressure of 7 X lom5Torr, Fe+ and FeCsHs+ have disappeared in favor of the parent ion and the product ion a t m/e 307. Double resonance experiments’ unambiguously identify the charge-exchange reactions, eq 1 and 2, and the condensation reaction, eq 3, as occurring in this sysFe+
I
/
reaction occurring a t fairly low pressure, but several other examples of this behavior have been noted r e ~ e n t l y . ~No ~.’~ protonated ferrocene is observed in pure Fe(CsH5)2, a result discussed further below. Recently developed ICR trapped-ion techniques* were applied to this system to accurately determine the rate constants of reactions 1-3. The variation with time of the four ions observed in ferrocene following a 10 msec, 70 eV electron beam pulse is shown in Figure 1. As expected from the reaction scheme, the Fe+ and FeCsH5+ ion intensities decay exponentially with time. From the slopes of the lines in Figure 1 and the known pressure, rate constants for the disappearance of Fe+ and FeCsHs+ are determined to be 9.8 f 1.0 X and 7.9 f 0.8 X cm3 molecule-’ sec-I. The former is identified with kl and the latter with k2 k3. The exponential rise of Fe2(C~Hs)3+(see Figure 1 ) provides a n independent determination of k3, which is measured as 1.2 f 0.1 X cm3 molecule-] sec-’. Subtracting this from the total rate for disappearance of FeCsH5+ gives k2 = 6.7 f 0.7 X cm3 molecule-’ sec-I. Table I lists the ion-molecule reactions occurring in ferrocene along with the rate constants measured in this work, those reported by S c h i l d c r o ~ tand , ~ those predicted by the Langevin polarization theory.14 The absolute rates determined by high-pressure mass spectrometry are about three times those determined by ICR, a rather poor agreement between the two very different techniques. However, the ratios between the two sets of data are constant within experimental error (k/ks in Table I). Correspondingly, the relative rate constants in each set of data are also identical within experimental error. The discrepancy between the two sets of rate determinations can thus be attributed entirely to + uncertainties ,~~ in the absolute pressure. The indirect pressure measurement technique used by Schildcrout4 involves a n ionization gauge far removed from the source region of the mass spectrometer and at lo4 lower pressure. As he points out, the method is subject to considerable uncertainty. The
+
Ion-Molecule Reactions and Gas-Phase Basicity of Ferrocene
4816 Table 1. Ion-Molecule Reactions and Rate Constantsa in Ferrocene
Reaction
kb
kSC
k/ks
Fe+ + Fe(C,H,), -+Fe(C,H,),+ + Fe FeC,H,+ + Fe(C,H,), + Fe(C,H,),+++ FeC5H, FeC,H,+ + Fe(C,H,), Fe,(C,H,),
0.98 f 0.10 0.67 i 0.07 0.12 * 0.01
2 6 ? 0.5 2.1 f 0.7 0.34 i 0.12
0.38 ?: 0.11
1.55
0.35 f 0.16
1.19
kLd
a All rate constants in units of 10- cm3 molecule-' sec-'. b This work. C Reference 4. d Langevin polarization theory rate; see ref 14. The molecular polarizability of ferrocene isgiven as 18.9 X lo-'' cm3 in Ya. G. Dorfman, Russ. J. Plzys. Chem. (Engl. Transl.), 37, 1347 (1963).
Table 11. Proton Transfer Reactions Involvina Ferrocene
Reaction observeda
H,S+ + Fe(C,H, j, PH,+ + Fe(C,H,), NH,+ + Fe(C,H,), (i-Pr),OH+ + Fe(C,H,), CH,NH=NH,+ + Fe(C5H,) HFe(C,H,),+ + CH,NH,
Thermochemical inferenceb
HFe(C,H,),+ + H,S HFe(C,H,),+ + PH, + HFe(C,H,),+ + NH, HFe(C,H,),f + (i-Pr)20 HFe(C,H,), + CH,N=NCH, + CH,NH,+ + Fe(C,H,), -+
+
-. -+
PA[Fe(C,H,),] 2 170C PA[Fe(C,H,),] B 188d PA[Fe(C,H,),] B 2 0 7 C PA[Fe(C,H,),] B 208e PA[Fe(C,H,),] B 209f PA[Fe(C,H,),] < 216g
a dk/dE negative in each case. Reverse reaction not observed. b Values in kcal/mol. PA = proton affinity, defined in text. C M. A. Haney and J. L. Franklin, J. Phus. Chem., 73,4328 (1969). dReference 12. e M . Taagepera and R. W. Taft, unpublished results. f M . S. Foster and J. L. Beauchamp, Int. J. Mass, Spectrom. Ion Phys.,15,429 (1974). gReference 23.
direct method used in the present ICR experiments has proven highly accurate in previous work and is expected to yield absolute pressures to within f 1 0 % in this situation. The total ICR reaction rates for Fe+ and FeCsHs+ are about two-thirds of those predicted by the Langevin formulation. Those of Schildcrout are a factor of 2 higher. Rarely do measured rate constants exceed the Langevin rate in nonpolar systems. Basicity of Ferrocene. Mixtures of ferrocene with select molecules were examined to delineate the nature of proton transfer reactions occurring in these systems. Molecules examined included H2S, PH3, "3, (i-Pr)zO, CH3N-NCH3, and CH3NH2. Proton transfer reactions 4-8, confirmed in each case by double resonance, were observed in binary mixtures of ferrocene with the respective neutral molecule. In a mixture of Fe(CsH& with a large
basicity for ferrocene in solution.] This has been interpreted as reflecting the rather nonspecific diffuse iron localized nature of the e2g ~ r b i t a l . ' In ~ ,contrast, ~~ the gas-phase results show ferrocene to be surprisingly basic, with a proton affinity higher than that of ammonia (eq 6). The metalbased electrons in ferrocene thus have substantial donor capabilities, consistent with the proposal that the metal atom is the site of attack in electrophilic substitution reactions.2,6 A similar conclusion is appropriate in the case of iron pentacarbonyl, whose proton affinity has recently been measured as 204 f 3 kcal/mol.21 This molecule also protonates on the metal atom in solution.22 The significantly higher basicity of Fe(CsH5)z relative to Fe(C0)s may reflect the fact that the metal d electrons in the latter are more involved in retrodative K bonding to the C O groups and thus reduce the electron donor ability of the metal relative to ferrocene. The tremendous difference observed beH3S+ + Fe(CsHs)2 HFe(CsH&+ + H2S (4) tween the solution and gas-phase basicity of ferrocene probably reflects the understandably poor solvation of the proPH4+ + Fe(CsHs)2 HFe(CsHs)2+ + PH3 (5) tonated molecule. NH4+ + Fe(CsH5)z HFe(CsH&+ + NH3 (6) The homolytic bond dissociation energy D(M+-H), deM+ fined as the enthalpy change for the reaction M H + (i-Pr)zOH+ + Fe(CsH5)2 HFe(CsH&+ + (i-Pr)zO (7) H , is related to the proton affinity of M by eq 10. This CH,NH==NCHj+ + Fe(CsH& HFe(CsH&+ + CH3N=NCH3 PA(M) - D(M+-H) = IP(H) - IP(M) (10) (8) dissociation energy is useful in correlating gas phase basiciexcess of CH3NH2, no protonated ferrocene is observed alties because it is generally constant for a homologous series though CH3NH3+ is the dominant ion present. But in a of molecules, leading to a linear relationship between 2:1:2 mixture of Fe(CsH&, CH3NH2, and HIS, reaction 4 PA(M) and IP(M).7,23From the ionization potential of feroccurs readily and is followed by reaction 9 in which rocene (6.8 f 0.1 eV),2aD[H-Fe(C5H5)2+] is calculated to HFe(C5H5)2+ transfers a proton to CH3NH2. The proton be 56 f 6 kcal/mol. This number is extraordinarily low compared with those for other organic and inorganic moleHFe(CsHs)2+ + CH3NH2 CH3NH3+ + Fe(CsH& (9) cules (which generally range between 80 and 130 kcal/ mol)7 and accounts for the absence of HFe(C5H5)2+ with affinity of a molecule M , PA(M), is defined as the enthalpy ferrocene alone in the gas phase. The analogous bond enerchange for the gas-phase reaction M H + M H + and represents a quantitative measure of intrinsic b a ~ i c i t y . ~ , ' ~gy in HFe(C0)5+ is calculated to be 74 f 3 kcal/m01,~~ Thermochemical data relevant to reactions 4-9 are presentstill low by comparison with other molecules but substaned in Table 11, from which the proton affinity of ferrocene is tially higher than that in ferrocene. Not surprisingly, established as 213 f 4 kcal/mol, corresponding to Fe(CO)5 and Fe(C5Hs)2 cannot be considered two members of a homologous series from the standpoint of (M+-H) AHf" [HFe(CsH5)2+] = 205 f 5 kcal/mol.I6 bond energies. The molecular orbital in ferrocene from which the lowest It is found in all molecules studied to date that (M+-H) energy ionization process occurs is the e2g orbital, localized bond energies are greater than (M-H) bond energies in the largely on the iron atom and consisting of the metal d, and corresponding isoelectronic neutral molecule [e.g., d,2-,,2 Presumably, protonation of ferrocene D(PH3+-H) > D(SiH3-H)].7,25 Such a situation may obalso involves this orbital, which is consistent with the welltain for transition metal-hydrogen bonds as well; the accepted observation that ferrocene protonates in solution (Mn-H) bond energy in H M n ( C 0 ) s is low, certainly less on the metal Protonation only occurs in the presthan 74 kcal/mo1.26 In view of the already weak [Hence of strong acids, however, suggesting a relatively low
--
7v20
-+
-
+
-
Journal of the American Chemical Society
/
+
97:17 j August 20, 1975
-
4817 Fe(CsHs)2+] bond, that in HMn(C5H5)2 may be especially so, perhaps accounting for the fact that the latter species has never been observed while its Tc and Re analogs are well-known. The extension of these results to other organometallic complexes, containing a variety of metal atoms and ligands, should prove illuminating. Acknowledgment. This research was supported in part by the United States Atomic Energy Commission under Grant No. AT(05-3) 767-8 and by the National Science Foundation under Grant No. NSF-GP-18383. References and Notes (1) Dreyfus Teacher-Scholar, 1971-1976. (2) For a review of metallocenes, see, for example, M. Rosenblum, "Chemistry of Organometallic Compounds", Wiley-Interscience. New York, N.Y.. 1965, Part I. (3) Previous mass spectral results are summarized in (a) G. M. Begun and R . N. Compton, J. Chem. Phys., 58, 2271 (1973); (b) R. W. Kiser in "Characterization of Organometalllc Compounds", M. Tsutsui, Ed., Wiley-Interscience, New York, N.Y., 1969, Chapter 4; (c) S. Evans, M. L. H. Green, B. Jewitt, A. F. Orchard, and C. F. Pygall, J. Chem. SOC., Faraday Trans. 2, 1847 (1972). (4) S. M. Schildcrout, J. Am. Chem. SOC.,95, 3846 (1973). (5) M. S. Foster and J. L. Beauchamp. J. Am. Chem. SOC., 93, 4924 (1971); preceding paper in this issue. (6) For a recent discussion, see D. W. Slocum and C. R. Ernst, Organornet. Chem. Rev., Sect. A, 6, 337 (1970). (7) J. L. Beauchamp, Annu. Rev. Phys. Chem., 22, 517 (1971).
(8) T. B. McMahon and J. L. Beauchamp, Rev. Sci. Instrum., 43, 509 (1972). (9) T. 8. McMahon, M. S.Foster, and J. L. Beauchamp, to be published. (10) J. T. S. Andrews and E. F. Westrum. Jr.. J. Organomet. Chem., 17, 349 (1969). (11) R. J. Blint, 1.B. McMahon, and J. L. Beauchamp, J. Am. Chem. SOC., 96, 1269 (1974). (12) R. H. Staley and J. L. Beauchamp, J. Am. Chem. SOC.,06, 6252 (1974). (13) R. D. Bach, J. Gauglhofer, and L. Kevan, J. Am. Chem. Soc.,04, 6860 (1972). (14) G. Gioumousis and D. P. Stevenson, J. Chem. Phys., 29,294 (1958). , (15) E. M. Arnett, Acc. Chem. Res., 6. 404 (1973). (16) Calculated using AYo[Fe(CsH&] = 51.3 f 1.3 kcal/mol glven in J. D. Cox and G. Pilcher, "Thermochemistry of Organic and Organometallic Compounds", Academic Press, New York, N.Y.. 1970, p 492. (17) T. J. Curphey, J. 0. Santer. M. Rosenblum, and J. H. Richards. J. Am. Chem. SOC.,82, 5249 (1960). (18) I. Pavllk and J. Subrt, Collect. Czech. Chem. Comm., 32, 76 (1967). (19) J. C. Kotz and D. 0. Pedrotty, Organomet. Chem. Rev., Sect. A, 4, 479 (1969). (20) M. L. H. Green, J. A. McCleverty. L. Pratt, and G. Wilkinson, J. Chem. SOC.,4854 (1961). (21) M. S.Foster and J. L. Beauchamp. to be published. 122) A. Davlson. W. McFarlane, L. Pratt, and G. Wilkinson, J. Chem. Soc., 3653 (1962). W. G. Henderson, M. Taagepera, D. Holtz, R. 1.Mclver, Jr., J. L. Beauchamp, and R. W. Taft, J. Am. Chem. Soc.. 94,4728 (1972). Calculated from eq 10 using IP[Fe(C0)5] = 7.96 eV from 0. Distefano, J. Res. Natl. Bur. Stand., Sect. A, 74, 233 (1970). R. H. Wyatt, D. Holtz, T. B. McMahon, and J. L. Beauchamp, Inorg. Chem., 13, 1511 (1974). F. E. Saalfeld, M. V. McDowell, J. J. DeCorpo, A. D. Berry, and A. G. MacDiarmid, lnorg. Chem., 12, 48 (1973), and references therein.
Photochemical and Photophysical Processes in Acetophenone Michael Berger and Colin Steel* Contributionfrom the Department of Chemistry. Brandeis University, Waltham. Massachusetts 021 54. Received November 25, 1974
Abstract: Besides the first-order radiative (1.2 X lo2 sec-I) and radiationless (5.3 X IO2 sec-I) decay channels the lifetime and emission yield of gaseous triplet acetophenone depends upon (a) diffusion to followed by quenching at the wall, (b) selfquenching (8.9 X IO7 M-l sec-'), and (c) photoreaction. The contribution of diffusion can be treated quantitatively and it is shown that the wall deactivates with close to unit efficiency. At sufficiently low pressure following SO S2 excitation, triplet acetophenone dissociates with unit probability to give C6HsCO- CH3.. The dissociative state can be deactivated by added 275 nm. There is gas and both emission and chemical quantum yields give a dissociative lifetime of -5 X IO-* sec for X,i, evidence that these lifetimes are wavelength sensitive. SO SIexcitation does not result in the above reaction although there is evidence for a temperature-dependent decay.
-
I n a previous paper we discussed the photochemistry of benzaldehyde,' and in this paper we report on our studies of acetophenone.2 Benzaldehyde photochemistry is very sensitive to the energy of excitation and on the nature of the electronic state excited. For example SO S I excitation (n r*) resulted in free radical formation; SO Sz excitation (r r*) led to an intramolecular process resulting in benzene and C O without the formation of free radicals. The triplet yield was somewhat wavelength dependent, and it has been suggested that the rate of intersystem crossing decreases with an increase in excitation ~ a v e l e n g t hThus . ~ we were particularly interested to determine the nature of and the relative importance of the primary photochemical processes in gaseous acetophenone as a function of excitation energy. Furthermore a knowledge of these primary photochemical processes was necessary before studying the bimolecular reactions of triplet acetophenone in the gas phase. The results of these latter studies will be reported in another paper.4 Early s t u d i e ~ of ~ ?the ~ photolysis of acetophenone were directed t o the search for a convenient source of phenyl rad-
- -
icals in the gas phase. Both methyl and phenyl radicals were produced in sufficient quantity for rate studies when photolyses using an unfiltered mercury arc were carried out at elevated temperatures ( >280°).6 Carbon monoxide, benzene, ethane, methane, and trace amounts of biphenyl and toluene were produced under these conditions. The use of acetophenone as a radical source in the glass was also attempted.7 However, full mercury arc irradiation of acetophenone i n a 5: 1 isopentane-3-methylpentane glass at liquid nitrogen temperature resulted in no detectable decomposition. The above mentioned studies reveal little about the nature of the primary photochemical processes involved since unfiltered light was used. Smith and Calvert8 studied the SO S I photolysis of trifluoroacetophenone at 366 and 3 1 3 nm at elevated temperatures (1 50-300O). They concluded the quantum yield for decomposition was extremely small, under their conditions: for example, at 366 nm and 280'. ~ c F , . 0.008.
-
+
Berger, Steel
-
+
+
-
Experimental Section ( 1 ) Materials and Apparatus. Acetophenone was distilled on a
/
Photochemical and Photophysical Processes in Acetophenone
